Duverger’s Law, and why third parties in the US benefit the one they disagree with most

I often get the question from (continental) Europeans and Israelis, “why does the US only have two parties?” Sure enough, it has more than two, but even the Libertarian party, for better or for worse, is barely even a niche player.

How come? A French political scientist named Maurice Duverger (who himself had a very “colorful” political career, to put it mildly) made the empirical observation that:

[T]he simple-majority single-ballot system [a.k.a., plurality voting, a.k.a. first-past-the-post] favours the two-party system.

Duverger, Maurice (1964). Political parties: their organization and activity in the modern state. Internet Archive. London : Methuen. p. 217

as well as its corollary

[B]oth the simple-majority system with second ballot [i.e., runoff voting] and proportional representation favour multi-partism.

ibid, p. 239

This is generally referred to as “Duverger’s Law”, but really is an empirical observation rather than a mathematical law (the way characteristics of some electoral systems are subject to mathematical proof, e.g. whether or not they satisfy the Condorcet criterion).

A generalization of Duverger’s law I have read is that “the number of viable political parties equals one plus the number of seats per constituency”. In the extreme case where the whole country is a single constituency (i.e., pure proportional representation), you can theoretically get as many parties as you have seats in parliament — Israel was not that far from this chaotic situation until an electoral threshold was introduced.

OK, you say, the US has one Representative per congressional district, and states in Presidential elections are winner-takes-all with the exception of Nebraska and Maine. And yes, in the 1992 US Presidential election, Ross Perot pulled 18.9% but not a single electoral vote.

But what about other countries with plurality voting? How come they have more parties?

Well, consider the UK. For the longest time, it has only had two major parties — the Tories [a.k.a. the Conservative and Unionist Party] and Labour. In the 2015 general election, UKIP (the UK Independence Party) got 12.6% of the vote, but only a single Member of Parliament. (They actually did much better in the 2014 European Parliament elections — ironically, for a Euroskeptic party.)

Likewise, the Liberal Democrat representation in the House is not proportionate to its electoral following, with (2019 election) a measly 11 MPs for 11.6% of the vote. But the Scottish National Party punches way above its weight, translating just 3.9% of the national vote into a whopping 48 MPs (out of a total of 59 from Scots districts). Why? Because it is regionally dominant — it is effectively the first party in a region of the United Kingdom, and functionally nonexistent outside. Likewise, while Green Party’s 2.7% got it one (1) MP, Plaid Cymru [=Welsh nationalists] got 4 MPs (all in Wales) with just 0.5% nationally (but over 20% in Wales).

Electoral Reform Society (UK) via Time magazine. They consider this a bug of plurality voting, while others might consider it a feature.

The situation in Canada is somewhat similar, with the Conservatives and Liberals as the two major parties, and the Parti Québecois as the regionally dominant niche player, translating 7.63% into 32 seats (out of a total of 338), while the New Democratic Party got just 24 seats with 15.98% of the vote!

What does all the above mean in the US context?

A regional third party might be viable at the Congressional level, and might become a serious spoiler in a Presidential election. If tomorrow a politician with a rabid following (say, the 45th President) decided to found a Dixie Party, and it were to compete directly with the GOP in a cluster of Southern states, that hypothetical Dixie Party could become a force in Congress, and might even find itself wooed as a coalition partner under certain circumstances. But this would presuppose 1st party status in those states — the same number of votes spread in penny packets over the whole USA will get you zilch.

And guess who, in this scenario, would be laughing all the way to the bank in the presidential election?

A hypothetical “Patriot Party”, “MAGA Party”, or similar would largely compete for the same voters as the GOP. If it can drive the GOP out of the market in certain constituencies, it can become a player at the Congressional level; if it can make the GOP go the way of the Whig Party at the national level, it will effectively replace the GOP as the opponent of the DINOcrat Party. Anything less than these two scenarios simply strengthens the hand of the DINOcrats — whether we like this or not.

Consider the 1st round of the Senate elections in GA. Note: I am not saying you should, or shouldn’t, vote Libertarian: I am simply noting that if the Libertarian hadn’t run for the Perdue seat, Perdue would easily have made the 50% threshold needed to prevent a runoff election (which probably was ballot-stuffed by the DINOcrats), and we would have had GOP control of the Senate. Does anyone believe Senate control by the current version of the DINOcrat Party is somehow closer to libertarian aims? I somehow didn’t think libertarians subscribed to Lenin’s “worse is better”… (Yes, I know, Lenin referred to setting the stage for armed revolution.)

Seriously, as one who grew up in a modified Condorcet system and lives in a proportional representation system, I am 100% sympathetic to the desire for a viable third party. All I am saying is: in a plurality voting system, pretty much the only kind of third party that can have any impact is a regionally dominant one — anything else in practice hurts the major party that is closest in views to the third party, as they would share a voter reservoir.

Finally, Donald Trump may have a number of flaws, but he wouldn’t have gotten as far as he did in business without being shrewd and understanding markets — electorates are a different kind of market, if you like. I would imagine that the above calculus (by him or one of his advisors) figured into his decision not to establish a new Patriot Party ticket.

3 thoughts on “Duverger’s Law, and why third parties in the US benefit the one they disagree with most

  1. Reblogged this on Head Noises and commented:
    Takeaway:
    Finally, Donald Trump may have a number of flaws, but he wouldn’t have gotten as far as he did in business without being shrewd and understanding markets — electorates are a different kind of market, if you like. I would imagine that the above calculus (by him or one of his advisors) figured into his decision not to establish a new Patriot Party ticket.

  2. While the US is currently on a 2-party system and will likely stay that way (or, if Dems have their way, perhaps reduce to a single party nation) it is still the case that one of the major parties will sometimes decay, die and be replaced by a new “major” party.
    The most recent example was the way in which the Whig Party was gradually replaced by the Republican Party. That change was fraught with danger. In fact, a brutal Civil War resulted. But there was a large group of committed Christians who fervently opposed slavery. They populated the new Repub Party and drove it to majority status.
    Someone more erudite than myself would need to comment on today’s situation and compare it to historic parallels. I only want to note that although all your points are sensible the current GOPe leadership has shown itself to only be willing to fight hard … against the wishes of their conservative base. That’s potentially an unstable situation.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s